[
https://issues.apache.org/jira/browse/NIFI-6428?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16883631#comment-16883631
]
Koji Kawamura commented on NIFI-6428:
-------------------------------------
Hi [~mattyb149], do you have any idea on how we can address this? My idea is
resetting 'inTransaction' to false when the processor restarts consuming binlog
with user defined initial binlog position (instead of managed state). Also, add
a description to note that specifying initial binlog position in the middle of
a transaction (between a BEGIN and a COMMIT) is not supported. How do you think?
> CaptureChangeMySQL throws IOException with BIGIN event due to lingering
> 'inTransaction' instance variable
> ---------------------------------------------------------------------------------------------------------
>
> Key: NIFI-6428
> URL: https://issues.apache.org/jira/browse/NIFI-6428
> Project: Apache NiFi
> Issue Type: Bug
> Reporter: Koji Kawamura
> Priority: Major
>
> Because of not clearing 'inTransaction' instance variable when the processor
> stops, when a user clears processor state and restart it with a specific
> initial binlog position, then if BEGIN event is received, the processor
> throws an IOException.
> The processor should reset 'inTransaction' to false, and also other instance
> variables when it start with an empty processor state.
> The issue was reported at NiFi user ML.
> [https://mail-archives.apache.org/mod_mbox/nifi-users/201907.mbox/%3C2019070919385393266224%40geekplus.com.cn%3E]
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
